Paper Procedure Kirkpatrick | Intervention training Duration Number (N) | Control training Allocation of controls Number (N) | Principal findings | Comments and reflections |
---|---|---|---|---|
Level 5 | ||||
Burden [66] CVC Level 4, 5 | ML: Didactic lecture and simulation practice with feedback 4 h N =? | Insertion in patients supervised by senior staff Historical controls N =? | Annual savings from decrease in infection $540,000 | Cost-benefit from actual financial data adds strength to conclusion. CBE as part of bundle, pre-post CBE-setup. Other interventions than only ML-course |
Cohen [67] CVC Level 4, 5 | ML: Lecture and simulation training with feedback 4 h N = 69 | Traditionally trained, five supervised insertions Historical controls N =? | Annual savings from decrease in infection $700,000 | Even including one-time investments, still resulted in positive Cost-benefit |
Sherertz [68] CVC Level 3, 4, 5 | CBE: Lecture, series of hands-on stations, one CVC. 3 h N = 406 | Conventional bedside and didactic instruction Historical controls N = 140 | Cost savings from decrease in infection 63,000-$800,000 | CVC Infection control course, large groups, other relevant procedures taught |
Level 4 | ||||
Evans [76] CVC Level 3, 4 | ML: Added lecture, video presentations, observed simulated hands-on 1–8 attempts N = 90 | “see one, do one, teach one.” Five supervised insertions on patients Concurrent, randomized controls N = 95 | Significantly higher first pass success rate in clinical setting | Ultra sound guided, Very low complications, pre and post-intervention |
Smith [81] CVC Level 3, 4 | CBE: Added case based didactic discussion, hands on simulation training 2 h N = 25(20) | Supervised performance on patients with immediate feedback Concurrent, randomized controls N = 27(8) | Intervention significant better knowledge and comfort in post-test, no difference to controls at 3 months. No difference in complications, nor needle passes | Skills decline over 3 months as argument for renewed skills training. |
Khouli [69] CVC Level 4 | CBE: Video and debriefing of hands on simulation training ? duration N = 24 | Historical: Apprenticeship model “see one, do one, teach one.” Concurrent: Video only Historical and Concurrent, randomized controls N = 23 | Significantly lower infection rate in interventional department than in the control group and historically | Strength from RCT-setup and well-defined control group training. Comparison to other enhanced training to account for Hawthorne effect |
Miranda [93] CVC Level 3, 4 | CBE: Presentation, observed and supervised hands on simulated training 2.5 h N = 40 | Usual ward orientation Firm based allocation N = 110 | Significantly larger increase in knowledge in intervention group, no difference in success rate. | No change in behavior or patient care, despite practical intervention. Infrequent insertion rate perhaps responsible for non-sustainable results. |
Udani [77] Spinal anesthesia Level 3, 4 | ML: additional training with deliberate practice and immediate feedback ? duration N = 10 | Base curriculum of written teaching materials and 15-min video Randomized, concurrent controls N = 11 | Significantly better checklist scores post-training, higher failure rate in intervention group | Randomized and well-described control group training. Immediate transfer of simulated best clinical practice skills to real patients, enhances patient safety in early procedural training |
Britt [70] CVC 3, 4 | ML: additional hand-on demonstration and performance ? duration N = 13 | Standard lecture followed by supervised training on patients Randomized, concurrent controls N = 21 | Nonsignificant lower complication rate in intervention group, no effect on infection rate | Randomized. Just short of statistical significance for level 4-measures, population too small. |
Barsuk [78] CVC Level 3, 4 | ML: lecture, ultrasound and simulator training with feedback 4–5 h N = 28 | Traditionally trained, five supervised insertions Historical controls N = 13 | Intervention group needed fewer needle passes in clinical performance | Only self-reported data on needle passes and self-confidence, introduces possible reporting bias. |
Barsuk [71] CVC Level 4 | ML: lecture, ultrasound and simulator training with feedback 4–5 h N = 92 | Lecture series, no formal training Historical controls in same ICU and concurrent controls in other ICU N =? | Significantly lower infection rate in intervention group compared with historically and control group | Historical controls, no control for Hawthorne effect of altered behavior not stemming from the practical hands-on training. |
Peltan [94] CVC Level 3, 4 | ML: added supervised practice on simulator 1–2 h additional N = 36 | Lecture, interactive online module, familiarization to CVC equipment, instruction at all procedures Randomized, concurrent controls N = 37 | Significant improvement in adherence to procedural protocol, no difference in other clinical performance | Strength from randomization. Direct observation enhances reliability of results. Equal clinical performance raises questions of the appropriateness of procedural protocol for improving outcome. |
Barsuk [72] CVC Level 3, 4 | ML: lecture, ultrasound and simulator training with feedback 4–5 h N = 51 | Lectures and by observing more experienced physicians performing CVC Historical controls N =? | Significant decrease in infection rate post-intervention in different hospital setting. | Enhances generalizability for results of the intervention, highlights the effort needed for implementation to succeed and the vulnerability of the intervention. |
Sekiguchi [73] CVC Level 3, 4 | CBE: Interactive video, hands on training 105 min N = 56 | Supervision of 10 subclavian, 10 internal jugular and 5 femoral vein insertions or 10 ultrasound guided procedures Historical controls N =? | Significant post-interventional decrease in complications, interns as fellows and attending physicians | Coinciding with increase in Ultrasound Guided insertion, which in itself decreases risk of mechanical error, possible confounder. |
Hoskote [74] CVC Level 4 | CBE: Debriefing on simulated pre-test, hands on training and repetitive simulated practice ? duration N = 60 | Not stated Historical controls N =? | Change in policy owing to decrease in infection rate to benchmark level | Good example of organizational change due to enhances in procedural safety following CBE training program |
Koh [83] CVC Level 3, 4 | CBE: lecture, video demonstration, simulation ? duration, 5 CVCs N = 32 | No controls | Learning curve of 7 CVCs performed before acceptable complication and success rate reached | Not directly related to the training course, but interesting to establish learning curve |
Martin [75] CVC Level 3, 4 | CBE: Didactic sessions, supervised skills training on cadaver, videotaped and reviewed for repeated sessions ? duration N =? | Advanced cardiac life support and advanced trauma and life support courses Historical controls N =? | Significant decrease in pneumothorax at 3 months, non-significant at 1 year | Although pre-graduate intervention, the results are postgraduate. Argument for effect of early training despite many procedures trained at the same time |
Smith [109] Fiberoptic intubation Level 4 | ML: Written instruction, simulated then clinical supervised training Min 1 h N = 12 | No controls | 95% completed within benchmark duration | Learning curve interesting for expected skills development in training. |
Barsuk [99] CVC Level 4 | ML: lecture, ultrasound and simulator training with feedback 0 h N = 102 | Lecture series, no formal training Historical controls N =? | Trickle-down effect of pre-test increase after first years of mastery learning course | Potential for additional effect of a training program, what kind of learning is transferred passively. |
Level 3 | ||||
Friedman [95] EDC Level 3 | CBE: Lectures on EDC insertion. High fidelity EDC-model 60 min, 15 insertions N = 12 | CBE: Lectures on EDC insertion. Low-fidelity banana model Randomized, concurrent controls N = 12 | No difference between hi- and lo-fi intervention, higher score by experience | Effects from inexpensive models comparable to more expensive could reduce costs of training, leading to higher cost-benefit |
Scavone [80] General Anesthesia Level 3 | CBE: General anesthesia for emergency cesarean delivery ? duration N =? | Lecture and General anesthesia scenario, unrelated to obstetric emergency Randomized, concurrent controls N =? | No difference in time to incision or confidence, I-group better score in repeated simulation | Adherence to scoring system perhaps enhanced safety, but did not lead to earlier operation, which would be a desirable outcome in real world. |
Gaies [85] Bag mask ventilation Level 3 | CBE: didactic session, observation and hands on, supervised practice ? duration N = 18 | Observing more experienced clinicians Block randomized controls N = 20 | Significant skills decline in both groups in final test | Rapid decline in skills after early skills training, rarely performed procedure |
Kulcsar [98] Spinal Anesthesia Level 3 | CBE: Same teaching, but by simulator with haptic feedback 110 min N = 14 | Practical procedural subparts teaching, using an orange Randomized, concurrent controls N = 13 | Non-significantly increased scores on clinical performance. | Less than half were tested clinically, very small study groups. Short follow-up 3 weeks. |
Barsuk [96] CVC Level 3 | ML: lecture, ultrasound and simulator training with feedback 4–5 h N = 76 | Lecture series, no formal training Historical Controls N = 27 | No difference in quality indicators in clinical performance between groups | Only self-reported data of complications, risk of reporting bias in the intervention group. |
Chan [101] CVC Level 3 | CBE: Instruction and demonstration in parts, followed by practice 61/50 min average N = 11 (part) | Instruction and demonstration in whole procedure, followed by practice Randomized, concurrent controls N = 8 | Only Part Task significant better in Global Rating Scale at one-month retention, rest no difference. | Interesting that the difference was found in the overall global rating scale and not in the check lists for parts of the procedure, when comparing whole to part-task instruction |
Friedman [79] EDC Level 3 | CBE: Added 17 min demonstration video on aseptic technique 75 min N = 18 | Lecture on aseptic technique Historical controls N = 11 | Significantly better scores at all intervals and in overall score of skills retention | Unclear to the extent of difference in training, only a new video or the subsequent clinical supervision as well? |
Ortner [84] General Anesthesia Level 3 | CBE: Full-scale general anesthesia scenario, supervised and debriefed ? duration N = 24 | Traditionally trained attending physicians as benchmark performers N = 6 | Trainees reached benchmark level of attending physicians immediately and at 8 months | The short course seems as effective as experience in sustainable skills for a multidisciplinary procedure |
Finan [97] Airway management Level 3 | CBE: Didactic component, demonstration and supervised hands-on simulator training 2 h N = 13 | Standard course, one of more skills training sessions and subsequent clinical experience Historical controls N =? | Significant lower clinical success rate and return to baseline skills after immediate effect | Kirkpatrick level 1 and 2 reached but could not be transferred into clinical practice. Cause? Fidelity, simulation not encompassing the variability of real life? |
Millington [82] CVC Level 3 | CBE: Multimedia educational material, demonstration followed by hands-on training 2 h N = 30, 16 in retention phase | No controls | Significant increase in retention of knowledge, immediate post-training increase in skills also | Retention test of skills would have been preferable to knowledge retention as an effect measure since other studies have shown retention discrepancies between the two. |
Garood [86] CVC Level 3 | CBE: One of more skills training stations in one day course. Small group training One day N = 41 | No controls | Immediate confidence increase, significant decrease at 3 months | Self-reported confidence in clinical encounters is a weak measure of learning effect, subject to reporting bias. |
Lenchus [100] CVC Level 3 | CBE: Video instruction, discussion, instruction on ultrasound guidance, demonstration, individual practice 160 min N = 60 | No controls | Significant improvement in clinical performance score | Very short training time, but until competency? Better adherence to checklist = better procedural performance or clinical outcome. |
Lenchus [103] CVC Level 3 | CBE: Video instruction, discussion, instruction on ultrasound guidance, demonstration, individual practice 4 h N = 60 | No controls | Significant immediate improvement in knowledge and procedural checklist score | Same setting as above, unclear if the post-instruction score was on the first real patient performance. |
Thomas [87] CVC Level 3 | ML: Instructional video, supervised hands-on training 60–90 min N = 26 | No controls | Confidence significantly improved at 3 months, clinical scores deteriorated. | Argument for mandatory retesting and training, as residents own perception of skills was incongruent with actual skills performance after three months. |
Barsuk [88] CVC Level 3 | ML: lecture, ultrasound and simulator training with feedback 4 h N = 49 | No controls | Significant decline in skills test at 6 and 12 months after initial improvement. | Another powerful argument for repeated testing and remedial training, skills decay over time if not. |
Laack [89] CVC Level 3 | ML: interactive learning stations of part tasks, supervised 4 h N = 26 | No controls | Significant skill decay after 3 months | Remedial training argument for maintenance of initially acquired skills |
Siddiqui [90] EDC Level 3 | ML: Lecture, video, hands-on training on lo-fi model Duration > 45 min N = 21 | No controls | Retention score consistently over benchmark | Strong argument for hands-on training, also for aseptic technique |
Diederich [91] CVC Level 3 | ML: Low-fidelity mannequin trainer, instructional video, partwise instruction and hands-on training with immediate feedback ? duration N = 20 (Low-fidelity) | ML: High-fidelity mannequin trainer, instructional video, partwise instruction and hands-on training with immediate feedback N = 20 (High-fidelity) | Both groups performed above the minimum passing score at 4 weeks retention test | Strength from randomization and from well-defined ML-interventions in both groups. Possible cost-saving potential from low-fi non-inferiority. Short follow-up (4 weeks). |
Cartier [92] CVC Level 3 | CBE: Instructional video and hands-on, videotaped simulations, supervised by peers ? duration N = 37, 18 sustainability tested | No controls | Significant skills and knowledge increase from pre-training to post training and subsequent decline to > 2 years sustainability test. | Possible Hawthorne-effect from one cohort pre-post testing. Large dropout to sustainability. Still effect of training although diminishing after 2 years as argument for remedial training at interval shorter intervals than 2 years. |